MLOps Engineer

About the role
This role is about turning fast-moving ML research into reproducible, reliable, and shippable systems without slowing down experimentation. You will own the infrastructure that connects experimental models to production, working across training pipelines, data infrastructure, evaluation, deployment, monitoring, and internal tooling.

This role is ideal for someone who enjoys the full lifecycle of machine learning systems: making research work in messy production environments, data useful, deployments robust, and creating feedback loops from real-world deployments so models can continuously improve.

Responsibilities

• Build and maintain training, evaluation, and deployment infrastructure for computer vision and spatial perception models.
• Develop systems for ingesting, organizing, validating, and learning from real-world video and sensor data.
• Create monitoring and diagnostics for model quality, runtime performance, data drift, and failure modes.
• Support efficient use of local and cloud compute for training, evaluation, and release workflows.
• Improve experiment tracking across research and engineering.
• Work with co-founders and teammates to turn promising models into production-ready components.
• Build internal tools that make it easier for the team to inspect data, compare experiments, and understand regressions.
• Contribute to engineering standards around CI/CD, testing, reliability, and release quality.

What we are looking for

• Strong Python skills and solid knowledge of software design
• Experience building production grade ML infrastructure, data pipelines, and model deployment workflows.
• Familiarity with PyTorch or similar ML frameworks.
• Experience deploying and optimizing models for edge and embedded targets, including runtime tooling such as CUDA and ONNX.
• Practical understanding of experiment tracking, dataset management, orchestration, CI/CD, and monitoring.
• Ability to work close to research while maintaining production discipline.
• Comfort with dynamic early-stage startup work, where responsibilities can evolve quickly.
• Bonus points for experience with computer vision, robotics, C++, distributed training, production data loops, or agentic infrastructure for orchestrating intelligent systems and workflows.